Machine Learning Seminar: Methoden und Anwendungen (Sommersemester 2022)
Dozent:
Prof. Dr. Christoph Lippert
(Digital Health - Machine Learning)
Allgemeine Information
- Semesterwochenstunden: 4
- ECTS: 6
- Benotet:
Ja
- Einschreibefrist: 01.04.2022 - 30.04.2022
- Prüfungszeitpunkt §9 (4) BAMA-O: 30.09.2022
- Lehrform: Vorlesung
- Belegungsart: Wahlpflichtmodul
- Lehrsprache: Englisch
- Maximale Teilnehmerzahl: 10
Studiengänge, Modulgruppen & Module
- ISAE: Internet, Security & Algorithm Engineering
- ISAE: Internet, Security & Algorithm Engineering
- SAMT: Software Architecture & Modeling Technology
- SAMT: Software Architecture & Modeling Technology
Beschreibung
The list of projects + slides can be found here. Please fill in your name, first, and second preference into the sheet and send a mail to Matthias to officially register. Deadline is Tuesday, April 26, end of day.
This seminar consists of semester-long research projects. The projects span topics from core machine learning research, such as generative models, uncertainty quantification, and interpretability; as well as applications in the biomedical and health sciences, such as epidemiological N-of-1 trials, genetics, and medical imaging. Students are expected to work closely with their individual supervisors (PhD students and PostDocs at the Digital Health - Machine Learning group), make substantial progress on their task, and give a presentation at the end of the semester. Especially successful projects may additionally lead to the publication of a scientific paper.
Students are required to have good coding skills (language will depend on the topic, but mostly Python and R) and have at least a basic understanding of modern machine learning, e.g. through the Deep Learning lecture at HPI or similar online courses.
If you have questions regarding the structure or if this seminar is appropriate for you, feel free to contact Matthias.
Voraussetzungen
Precise requirements differ between the different research projects. In all cases, basic skills in machine learning/deep learning and/or statistics are highly preferred. Ambitious students may take this seminar in parallel with the Introduction to Deep Learning course.
Leistungserfassung
Students will work on a project for the course, and the seminar will end with a short presentation and/or a short written report. Details to follow.
Termine
We will have a kick-off meeting on April 22nd, at 9:15 AM. After that, supervisors and students arrange regular individual meetings.
The kickoff will be preferred in person, room G1.E15/16, but if you can't be there in person, you can also join via zoom:
uni-potsdam.zoom.us/j/62449109879
Meeting ID: 624 4910 9879
Passcode: 28774524
Zurück